Brake switch. resistance loads etc.

[stag]

 When my mill did not furl and ran away in a gale, shorting out the 3 phase wires made no apparent difference. Since then I have moved the switch to the foot of the tower, instead of at the end of a 40yd cable run. It now works ,but the mill starts slowly in fairly light winds and will still spin up pretty fast as the wind picks up. A few days ago, I banged two nails into apiece of wood and stretched a steel spring between them. I clipped the DC wires onto either end of the spring when a stiff breeze was blowing. The AC brake switch was on while connecting it, and then released. There was a small amount of warmth in the spring,but I think it is a bit heavy and too much like a short in the DC. The rectifier also got warm, which it had noticeably not done before. Also ,this held the mill down much  more than an AC short, and this being at the end of the cable run. So it seems that shorting the DC is a more effective brake , but could this damage the rectifier,or is this an acceptable practice.?

[OperaHouse]

I think the difference is explained by added resistance.  The three phase AC wires coming down the tower are likely much smaller than the DC line.  It is also possible that the wire used in the moll is smaller than it should be.  The result being that there is far too much resistance and there is too much blade for the mill.

[Flux]

The whole concept of braking is difficult to understand.

I suspect you have a mill without enough magnet and copper to brake reliably anyway, in which case braking is likely to do more harm than letting it fly.

An alternator with lots of copper anfd magnet in proportional to the blade size will not start up in any wind condition with the ac leads shorted. It will also almost invariably stop from any speed in any wind, perhaps rather more violently than ideally desirable.

If the alternator is nort stiff enough then it won't hold on an ac short and won't stop from all wind conditions ( if it stops at all).

This type of machine will at least slow down when loaded to the optimum point with resistance ( in the ac lines or in the dc side, doesn't matter).  If you can get it to slow to the point where you pull the prop way off its curve you may then be in a situation where it will finally brake on short circuit. In run away conditions this is worth a try but you need to get the process over and done with in a few seconds, otherwise it is safer to let it fly as there will be a lot of heat in the stator.

You obviously have a lot of line resistance after the rectifier, there is no idndication from your information whether you have too much resistance in the ac leads to the bridge or whether it is in the winding ( I suspect the winding).

You also seem to have an under rated rectifier or one with too little heat sink. I don't think this mill is a candidate for any form of electrical braking in its present state and persisting with the idea is likely to mean a cooked stator unless it is actuallr reactance limiting ( which we have no means of knowing without load test figures).

Flux

[stag]

perhaps I should point out that the 40yd cable run is 3 phase, and the rectifier is at the end of this,and not at the foot of the tower. The mill has been charging 12v batteries at 13/15v but when connected to the spring it dropped to 4v with much higher current. The stator is wound with 1.4mm wire and flux has previously suggested that it could be self limiting as it has run away in gales without burnout. it seems pretty safe charging12v and braking AC will keep it down a reasonable amount,so maybe I should continue in this mode. The heat sink is a large piece of aluminium channel with three ,600v 35 amp rectifier blocks. One for each phase with the wires split to use both AC terminals on each block. OH! and the 40yd, 3phase cable run is 6mm per cable. Thanks for your input guys. I keep on thinking and tweaking things but maybe I kinda got it right in the first place?

[Flux]

That does seem to imply that you have a fair bit of resistance in the cable run and it may be limiting your output under wiorking conditions. It seems as though you have a lot more current capability into a near short circuit than into the battery.

If it is not too difficult it may be well worth trying temporarily moving the batteries and rectifier to directly under the tower and see what difference you get in charging performance.

If you are getting enough to satisfy your nereds then fine, but I suspect there is more available. From previous descriptions it seemed as though the alternator was self limiting, in which case I wouldn't expect much more current ( rectifier heat ) into the low resistance of the spring. It may be that you are actually self limiting on blade speed and not the alternator.

Flux

[stag]

thanks flux. That's very helpful. If I  made a  coil of the right resistance,not approaching a dead short, might I be able to use it as a heater and get more power than battery charging? I guess this is what would be required for an effective dump load anyway?  I don't understand why increased resistance gave more current and less voltage. I will try what you suggest, and try battery charging at the foot of the tower. The simplest way seems to be take a couple of batteries down there and make another rectifier, as the rectifier blocks are pretty cheap.  If the blades are self limiting,it is at very high RPM. But now that I have it furling before 30mph ,it is still pretty fast but less scary. From your reply to my previous post,I gathered that I could expect much more power into a heating coil than into a   battery. Why is this, or did I misunderstand?

[Flux]

An ideal wind turbine will have blade speed rising directly with wind speed ( to maintain constant tsr). It will also have an input power from the blades that rises as the cube of the wind speed.

For battery charging with a highly efficient alternator the alternator speed will be clamped to something just over cut in speed and it will not let the prop give good power in any wind much above cut in.

To get better results you need to lower the alternator efficiency, you loose electrically but you gain more from the prop.

With a heater the power follows the square of the voltage ( and voltage is related to speed). In the perfect world this is still not the ideal cube law but in real life you can't have a 100% efficient alternator so you will find that the square law of the heater is in fact a close match to the prop. If you get things completely right with a heating load you can hold the prop near its peak output from near cut in to furling point. This gives a substantially greater output than the poorly matched case of the direct battery charger.

For large machines heating can be very effective, it is more difficult to justify it on small power set ups where the heat may be hard to use effectively.

50W of heat may not go far but the same power could run a laptop, charge mobile phones and in this day and age it can produce quite a useful lot of light. Small power machines are best used for high grade energy use.

I don't want to get too involved in this but although heating is basically simple and should be very efficient most people struggle with it because with a heater there is no direcrt equivalent to cut in of a battery charger and the prop is starting against a highly stalled condition. That means that a heater control needs some method of holding the heater off until the blades are producing sufficient power to do something useful. This is basically a simple problem to overcome but it seems to defeat most people. The other issue is to be able to get a good match to the blades from light wind to furling. If the alternator is very efficient you need something steeper than the square law and it becomes necessary to add additional heaters in parallel for best results.

Typically a good mppt battery charging machine can provide about 3 times the output of a battery charger optimised for low wind. This figure seems about right for mppt grid tie and I would expect the same to hold for a good heating set up.

Flux

[gww]

Stag
You might get better advice if you pointed out that your turbine is made with ferrit magnets.  I am sure on some advice it would make no differance but it gives one reason for the weakness of the shorting of the turbine.  Sometimes poeple don't remember which turbine they are speaking of as most are stronger due to neo's.  I am always glad to hear how it is working.  Did you ever get an amp meter?
Cheers
gww

[stag]

gww. I think you are exactly right. I do get the impression that,unless specified, people tend to assume that you are using neo magnets where ,I assume, most of the braking and stalling stuff comes from. Being a newbie I am rapidly learning that my project is something different. However! It's all good fun and I am about building something that works, with stuff I can get very cheaply or for free. recieved my amp meter from China a couple of days ago, and to my shame, haven't hooked it up yet. Will keep you posted. Flux.Thanks again for some excellent info. Would it make a difference if I put the rectifier at the foot of the tower and made the cable run DC. With the cable  I have I could have 2 cables of 9mm instead of 3 cables of 6mm. As for a relatively constant TSR , my blades speed up and slowdown significantly with changes in wind speed.

[Flux]

Not sure what cable you have to the rectifier but normally if you transmit dc instead of ac there is less cable loss.

If you only have 3 cables then the best you can do from the rectifier is run 2 in parallel on one polarity and one on the other, this does give less loss than running the 3 phase.

If your 3 phase is made up of more than 3 cables you may be able to balance things but the loss will be similar.

With ferrite magnets I would imagine you will never see anything approaching stall so the lower you can get the cable loss the better hence my suggestion to try the battery and rectifier at the tower. If the improvement is significant it will be worth finding more cable to go in parallel. Get your ammeter running then you can try these experiments more scientifically, looking at battery volts is not ideal.

Flux

[stag]

The cable is high temperature stuff, 3core/ 1.1/2mm per core. I twisted the cores together to give 3x1.1/2= 4.5mm on each phase. BUT,I had enough cable for one more run,SO I connected one of the 3 cores of this cable to each of the phases,giving 6mm on each phase. SO ,as I actually have 4 lengths of cable I could use 2 twisted together for each polarity DC.   2 cables with 6 cores twisted together would give 9mm.   This is splitting a total of 18mm between 2 DC cables or 18mm split between 3 AC cables. If this is the same or better than the present arrangement, changing to DC would at least place the rectifier at the foot of the tower,making it more convenient to make comparisons,as you suggest. Thanks again.

[DanG]

Me thinks your summing diameter is inserting an error - lucky mm² and diameter merge on this size..

You mentioned..."2 cables with 6 cores twisted together would give 9mm" which would be closest to 00 or 2/0 AWG.

But six 1.5mm diameter wires combine to yield a 7.5awg equivalent.

A handy rule of AWG is pairing two identical conductors yields a number three sizes lower.

So... 1.5MM is closest to 15awg...
if two 15awg's = 12awg (2 x 1.5)
if two 12awg's = 9awg   (4 x 1.5)
adding another 12awg yields 7.5awg. (6 x 1.5)
if two 9awg's = 6awg, so with eight 1.5mm's you'd have 6awg.

This could help explain the low current brake symptoms.

To get the 2/0 equivalent would require umnn.... 32, closer to 40 1.5mm wires combined...

[stag]

Dan G. I think I see my mistake.When combining 3x1.5mm squared cores, the sum of these is not squared,but just simple multiplication,thus introducing an error. Please correct me if I have misunderstood. As I have 4 lenghts of cable, I guess I should go ahead and use 2 cables(6 cores of 1.5mm) for each polarity DC until I can acquire more cable to beef things up. Thanks. Pete.

[stag]

A friend has suggested that pushing a given current through six thin wires could have more resistance than through one thicker wire , even if the total thickness is the same, bearing in mind that the thinner wires are only twisted together at the ends. If so, this could explain the losses when taking this approach?